Primary Liquefaction of Coal by Hydrogenation Chemical Nature and Effectiveness of the Vehicle C. H. FISHER AND ABNER EISNER
U.S. Bureau of Mines Experiment Station, Pittsburgh, Pa.
To obtain information on the relative effectiveness of various dispersion media in the primary liquefaction of coal by hydrogenation, Pittsburgh bed coal from the bureau's experimental mine was hydrogenated in over thirty vehicles under comparatively mild conditions (400" C. and 67 atmospheres original hydrogen pressure). Previous claims that tetrahydronaphthalene is a good dispersion medium were confirmed. Of the other hydrocarbons studied, only methylnaphthalene and diphenyl were of comparable effectiveness. Several completely aromatic
(naphthalene and chrysene) and saturated hydrocarbons (decahydronaphthalene and nhexadecane) proved to be poor vehicles. The presence of polar compounds, such as phenols and amines, in the vehicles was found to be beneficial. The products formed in highest yield were pitches, which melt at moderately elevated temperatures. Elementary analyses of these pitches indicated that hydrogenation had raised the carbon and hydrogen contents, lowered the sulfur and oxygen contents, and caused little change in the nitrogen content.
A
in phenol (23, 1 4 , and unsatisfactory results were obtained only with anthracite. I n one of the most recent and extensive studies of the effect of vehicle, no direct connection between the chemical nature and effectiveness of the vehicle was observed (IO). The presence of tar acids in the vehicle was shown to be beneficial, and a high-boiling oil from hydrogenated coal was found to be one of the best vehicles; it was surpassed only by a mixture of m-cresol and tetrahydronaphthalene. Warren (SO) found that high- and low-temperature tars, but not lubricating oil, are suitable for the hydrogenation of coal in a continuous system. To be completely satisfactory in coal liquefaction the vehicle must possess several important characteristics. The compound, or mixture of several compounds, must have high solvent and depolymerizing action on the coal. The numerous researches already carried out on the extraction of coal (2, 3) afford some basis for selecting solvents effective in this respect. In addition to dissolving the coal, the ideal vehicle should be a good solvent for hydrogen under the conditions of the operation. To remain in the liquid phase, the vehicle must have a critical temperature higher than that of the experiment. The melting point and viscosity should be comparatively low so that the solutions obtained can be easily filtered, centrifuged, etc. The boiling point, which must be as high as about 200' C. to afford a satisfactory critical temperature, should be low enough to permit easy separation of the vehicle from the hydrogenated coal by distillation. The ideal vehicle also should function as a hydrogen carrier. Permanent consumption of hydrogen by the vehicle is objectionable, but temporary absorption and subsequent transference of hydrogen to the coal substance appear to
LTHOUGH several important variables involved iii the liquefaction of coal by hydrogenation have received much attention, little work has been carried out to determine the importance and role of the vehicle. Primarily this is because vehicles other than oils obtained from coal by hydrogenation or carbonization have not been considered interesting from a practical standpoint. The possibility exists, however, that some individual compound, or mixture of several compounds, may be used economically in the initial stages to convert the coal into an ash-free material more suitable for further hydrogenation. A process of this kind, using a mixt.ure of tetrahydronaphthalene and phenols as vehicle, recently received considerable attention and reached the stage of The little evidence now commercial development (IS). available indicates that the nature of the vehicle is important, especially in the preliminary stages where the main purpose is to get the coal substance into solution. Therefore, a study of the vehicle's influence in coal hydrogenation is of practical as well as theoretical significance. In laboratory studies a few pure organic compounds have been used as vehicles. Beuschlein, Wright, and Williams (5) showed that phenoI or anthracene is preferable to diphenyl. Boomer and Saddington (7) found that the conversion obtained on hydrogenating various coals in phenol, tetrahydronaphthalene,, petrolatum, bitumen, and naphthalene was largely a function of the dispersion agent, tetrahydronaphthalene being the most effective. Pertierra (21, 22) and Pott and Broche (26) found that mixtures of tetrahydronaphthalene and tar phenols are better dispersion agents for coal hydrogenation under mild conditions than either tetrahydronaphthalene or phenols alone. Twenty-nine coals ranging in rank from lignite to anthracite were hydrogenated 939
INDUSTRIAL AND ENGINEERING CHEMISTRY
940
VOL. 29, NO. 8
TABLEI. PHYSICAL CONSTANTS OF VEHICLES Used in Expt. No. 39 9, 20 9, 20 9, 20 13 23 25 36 38 34 10
26 21 32 31 8 18 Q
b
Compound Used as Vehicle Acetamide n-Cresol o-Cresol
6:E1
Diphenyl Diphenylamine Diphenyl ether Diphenylmethane Glycol Naphthalene a-Naphthylamin e Pinene Quinoline Stearic acid Tetrahydronaphthalene o-Toluidine
Reference No. (16)
... ...
(1;. is, (9,'i 8)
(ii)
(11)
iih (4, 6 ) ... ...
(18) (29)
...
Melting Point
c.
81 11 31 35-6 69170 53 27 25 -15 80
50 -55 -19.5 69-70 -31 -16
Boiling Point a
c.
222 203 191 202 191-3 265 302 259 258 197 218 301 158-61 237 291 (100) 205-7 200
Critical Temp.
c. ...
432 422 426 384 528
... ... ... ... ... ...
>bib ...
421
...
Surface Tension at 200 C." 39.3 37.4 39.8 36.7 26.7 29.5 37.7
(85' C.) (10'C.) (15.5O C.) (129 O C.) (80' C . )
3 j : i (300 c.) 47.7 39.0'~ 51.1b 2 7 . 0 (10' C.) 45.0 32.65 34.25 (18.3' C.) 40.0
Dielectrio Constant at 20" C.5
Dipole Moment
6 0 . 3 (83OC.) 5 . 0 24O C.) 5 . 8 24OC.) 5 . 6 2.4OC.1 2.11 2.62'J 3 . 3 (52O C.) 3.69 2.56 41.2
3.72 1.60 1.44 1.64
I
... ...
2.7 8 . 9 (21'C.) 2 . 3 2 (67O C.) 2.66 6.0
0.0
0.0 1.3 1.05
0.0
2.2
0.0
1.44 1.10 2.25
...
0.52 1.3
Unless otherwise noted. Extrapolated.
TABLE11. ANALYSES OF BRUCETON COAL(PER CENTBY WEIGHT) facilitate the l i q u e Volatile Fixed In most cases the Moisture Matter Carbon Ash v e h i c l e s w e r e profaction. It has been postulated (7, 8) that Proximate 1.6 36.7 56.4 6.3 cured in a satisfactory tetrahydronaphthaUltimate: 13 C N 0 S Ash state of purity and 4 s received 5.3 77.6 1.6 7.6 1.6 6.3 used as such. Decalene acts as a hydroMoisture-free 6.2 78.8 1.6 6.4 1.6 6.4 gen carrier or donor Moisture, ash-free 5.6 84.2 1.7 6.8 1.7 ... hydro n a p h t h a l e n e Si02 Ah03 Fez04 Ti02 PlOs CaO MgO $ 0 3 Nan0 K20 (Eastman Kodak in the hydrogenation Ash analysis, % ' 45.4 24.1 19.9 1 . 0 0.2 3 . 6 0.6 3.5 0.6 0.9 Company) was andextraction ofcoal u n d e r pressure and washed with concenthat its effectiveness t r a t e d sulfuric acid as a vehicle is due to this behavior. The possibility of using and distilled. The cresol (mixture of isomers), tetrahydrovehicles that may liberate hydrogen during the experiment naphthalene, and a-methylnaphthalene were redistilled, only (such as decahydronaphthalene, tetrahydronaphthalene, and the middle fractions being retained. Technical grades of hydrogenated quinoline) is extremely interesting. dimethylnaphthalene, chrysene, xylenol, and abietic and Vehicles for coal licluefaction also should be noncorrosive, stearic acids were used. nontoxic, and stable to thermal treatment. From a practical Hydrogenation Apparatus and Procedure standpoint it is desirable that the vehicle be readily available, easily recovered, or produced during the hydrogenation, and The hydrogenation experiments were carried out in a conit should form with pulverized coal a paste capable of being verter (1200-cc. capacity) made of 18-8 stainless steel. The pumped without settling. converter rested on rollers which caused rotation (about In the present work the vehicles were selected with little retwenty-three turns per minute) on the horizontal axis. The closure was a modification of the self-sealing Bredtschneider gard for practical considerations. Cresol, xylenol, catechol, joint recently described (28). The seal was effected by quinoline, toluidine, naphthylamine, diphenylamine, thiocresol, stearic acid, glycol, and acetamide were studied because means of turned copper rings made from hard drawn tubing of their polar properties. Naphthalene, methylnaphthalene, The initial sealing was made by drawing up the head with set diphenyl, diphenylmethane, and n-hexadecane were used to screws in a spider web arrangement attached to the head. Heat was provided by an electric furnace, the temperature illustrate the effect of aromaticity or paraffinicity. Decahydronaphthalene, tetrahydronaphthalene, pinene, nicotine, being indicated by a thermocouple in a well bathed by the conabietic acid, and ethyl dihydroabietate were selected priverter contents. The difference between the temperature of the converter contents and that indicated by the encased thermomarily because of the possibility that they would function as couple was determined and found to be about 4" at 400' C. sources and carriers of hydrogen. Some of the compounds A few preliminary experiments were conducted to arrive used as such or as mixtures in the present work for the liquefaction of coal are listed in Table I, together with physical a t the standard conditions to be employed in the subsequent work. Using tetrahydronaphthalene as the vehicle, satisdata. Data for which references are missing were taken from factory results were obtained a t 1000 pounds per square inch International Critical Tables. (70.3 kg. per sq. om.) initial hydrogen pressure and 400' C. Coal and Vehicles when either 1 or 0.1 gram of stannous sulfide was present. The liquefaction of the coal was much less satisfactory when The same coal, obtained a t the Bureau of Mines experimental mine a t Bruceton, Pa., was used in all the experiments. the stannous sulfide was omitted; this catalyst was therefore This Pittsburgh bed coal is classified as a high-volatile A cokused in the later experiments. ing coal (1) and falls toward the low-rank boundary of this Some of the previous investigators of hydrogenation vehicles have preferred to work without catalysts, and evidence is class. Plastic range curves obtained with coal from the exa t hand to indicate that differences in the effectiveness of perimental mine, which show that the plastic state is first vehicles are more pronounced in the absence of catalysts (IO). reached a t about 388" C., were recently given (26). The proximate and ultimate analyses of the coal used in the In spite of this apparent tendency of catalysts to make vehicle present work are given in Table IT. The sulfur was found to effectiveness uniform, a catalyst was used in the present experiments to effect considerable liquefaction under mild conbe present mainly in the pyritic and organic forms (0.83 and 0.71 per cent, respectively), sulfate sulfur being present to the ditions and to make the operating conditions similar to those of actual practice. extent of only 0.06 per cent. The fusain content of PittsThe following is a description of the procedure adopted: burgh bed coal generally is 2 to 3 per cent (12).
AUGUST, 1937
INDUSTRIAL AND ENGINEERING CHEMISTRY
941
TABLE111. DATAON VARIOUSVEHICLES .4
B
C
D
--ProductsLiqui ds , Vehicle solids Gases No. Loss Centrifuged Grams Grams Qrame Grams Hydrocarbon Vehicles 2 8 Tetrahydronaphthalene 193.2 6.8 2.0 171.7 16: 44 607 Naphthalene 50% tetrahydronaphthalene 193.1 4.2 3.7 180.6 17 3.3 Met%ylnaphthaledeo 191.2 6.6 175.8 Diphenyl 23 192.1 ... 187.0 Diphenglmethane 3.6 2.0 38 195.4 182.3 194.3 Chrvsene 46 Naphthalene i g i .9 4.6 4:s iii:o 10.27 Decah ydronaphthaIene 194.6 1.0 185.3 5.4 13 60% Pinene, 50% tetrahydronaphthalene 189.8 21 6.0 179.7 6.2 Cetane 192.7 30 3.6 183.9 4.7 4 0 7 Decahydronaphthalene, 60% naph11 tgalene 192.4 3.4 5.2 Hvdrocarbon-Phenol Vehicles 24, 42 195,7 4.8 0.5 184.8 192.8 4.2 4.0 170.8 7, 9 193.3 5.9 0.8 184.1 28 20.41 194.4 6.0 1.6 181.5 29 3.9 4.6 178.0 33 5.1 2.8 180.8 Amine and Hydrocarbon-Amine Vehicles 50 Tetrahydronaphthalene 60% quinoline 193.9 4.3 2.8 19 32 183.1 15: 18 5 0 5 Tetrahydronaphthalen; 60 toluidine 193.3 4.2 3.5 180.7 37 Tetrahydronaphthalene: 60% nicotine 191.6 6.4 3.1 176.1 E p g e nylamine, 1926 3.7 ... 179.7 25 a-Naphthylamine 191.7 3.3 6.0 164.5 26 Miscellaneous Vehicles 22 75% Tetrahydronaphthalene, 25% p-thio187.4 cresol 6.7 6.9 179.0 39 50% Tetrahydronaphthalene, 50% acet189.3 11.5 amide 0.2 179.7 Abietic acid 180.4 17.0 14 3.6 138.3 Eth 1 dihydroabietate 172.8 20.6 36 7.6 163.6 50 Tetrahydronaphthalene 50 glycol 21.3 175 34 162.0 50fTetrahydronaphthalene: 50% stearic 31 acid 194.1 7.6 $0.7 188.8 3.4 Diphenyl ether 196.8 36 186.2 0.8 a 60 per cent alpha- and 40 per cent beta-methylnaphthalene. b Approximate.
...
...
...
A mixture of 1 gram of stannous sulfide (prepared in the usual manner from the chloride) and 100 grams of coal (200-mesh) was placed in the bomb, and 100 grams of the vehicle were added. After the head was attached and the bomb flushed successively with nitrogen and hydrogen, hydrogen was introduced until the ressure was 1000 pounds per square inch. After 0.5 to 1 hour Rad elapsed, the pressure was again brought to 1000 pounds, replacing the hydrogen which had dissolved in the reactants. The temperature of the converter was then brought to 400' C. a t the rate of approximately 2' per minute and maintained at 400' (*3") for 3 hours. After the converter had cooled overnight to room temperature, the gaseous contents were passed through a wet test meter, and an average sample was collected by mercury displacement. Gas samples were analyzed in the Orsat apparatus, Bureau of Mines type (Table VII). The liquid and solid contents of the converter were transferred to a weighed bottle and centrifuged for 1 to 2 hours at about 2800 r. p. m. The ash-free oil thus obtained was removed from the bottle, weighed, and distilled. The amount of oil obtained from the centrifuge, calculated as per cent of the material centrifuged, is given in Table I11 (column E). The products not removed manually from the bomb were washed out with acetone and added to the main residue in the centrifuge bottle. The bomb was weighed before and after the acetone washing on a Troemner bullion balance to determine the amount thus removed. The combined residues were washed several times with acetone a t room temperature and dried t o obtain the weight of acetone-insoluble material. The acetone residues were washed with benzene for 48 hours in Soxhlet extractors, dried in an oven a t about 100" C. for 24 hours (or in a vacuum at 100' for 8 hours), and weighed to ascertain the amount of benzene-insoluble material (column F, Table 111). Since the degree of liquefaction is probably best indicated by the quantity of benzene-insoluble material, a check on the accuracy of the manipulations leading to the isolation of this material was desired. This check was obtained by determining the ash content of the benzene-insoluble residue and calculating the amount of ash recovered. Usually about 6 of the 7.3 grams originally present (6.3 grams of ash and 1 gram of catalyst) was accounted for in the benzene-insoluble residues. The conversion of pure coal was then calculated on the basis of complete ash recovery; these data, which are in satisfactory agreement with the yields (column G) determined in the usual manner, are given in column H.
E
F
Oil from Centrifuge
Benzene Insol. Grama
%
%
..
94.5 93.2 92.2 89.8 82.4 81.2 78.7 76.7 74.0 72.3
89.7 87.8 88.1 88.8 80.3
30 48 55 60
14.9 16.1 17.0 19.2 25.8 26.9 29.2 31.0 33.4 34.9
..
36.4
70.7
...
77 84 79 74 42
93.7 93.7 91.9 91.6 86.2 80.0
91.3 88.3 89.0
66
16.7 15.7 17.3 17.6 22.4 28.0
81 78 79 49 57
14.8 16.3 17.8 18.0 20.6
94.7 93.0 91.3 91.1 88.2
91.8
% 78 61 79 41 56
...
68.2
.... ..
...
...
85.0
...
... ...
87.3 .
I
.
77
16.2
93.1
90.9
77
30 63
..
18.8 22.1 24.2 24.6
90.2 86.6 84.2 83.8
8i:i
56 44
29.7 30.0
78.1 77.8
... ... ... 76.0
The centrifuged oil was distilled through a Vigreux column t o recover the unchanged vehicle and separate any low-boiling material. The vehicles of moderately low boiling point (naphthalene, Tetralin, cresol, toluidine, etc.) could be easily separated from the high-boiling product in this manner. Great difficulty was experienced in distilling the high-boiling vehicles, either a t atmospheric pressure or in vacuum, because of their tendency to foam or run over. The distillation residue, the main product of the liquefaction process, was conveniently poured from the' flask while hot. In a few instances the products were analyzed by a modified procedure. The solid roducts obtained in experiments 10, 11. and 46 were extracted Sirectly with benzene for 3 to 4 days, the extent of liquefaction being judged by the weight of benzene. insoluble residue. The products obtained in experiments 14, 23, and 27 were converted into viscous oils by warming to 5060' C., diluted with 30 cc. of benzene or Tetralin, and centrifuged. The centrifuge residues were then washed with acetone and extracted for 48 hours with benzene in the usual manner.
Nature of Products As obtained from the converter, the products ranged from viscous oils to comparatively firm solids. Products smelling strongly of ammonia were obtained in the a-naphthylamine and acetamide experiments, possibly because of the production of ammonia or aliphatic amines from the vehicles. In most cases the products had the odor of hydrogen sulfide. Centrifuging separated the crude products into : 1. Residues consisting of ash, fusain, stannous sulfide, and other material resistant to hydrogenation, as well as some of the solution containing the vehicle and dissolved products. 2. Small quantity of water and a solution of vehicle and hydrogenated coal. This solution appeared to be homogeneous but exhibited definite colloidal properties upon ultramicroscopic examination.
After being separated from ash and residues by centrifuging and from vehicles by distillation, the products were firm pitches at room temperature. At moderately elevated tem-
INDUSTRIAL AND ENGINEERING CHEMISTRY
942
peratures (100" to 150" C.) these products were soft or fluid. Although little volatile a t reasonable temperatures, distillation of the pitches a t 5 mm. to 200" C. produced small yields (8 to 10 cc.) of oil. Ultimate analyses of the pitches are given in Table IV. Since the pitches contained small amounts of ash (about 0.5 per cent) the separation of heavy oil and ash by centrifuging was not complete. As shown in the following table, extraction with benzene in the Soxhlet apparatus caused about 70 per cent of the pitches to dissolve:
-
7
Pitch No.
24 hr.
12 63.3 15 36.4 17 22 50:6 32 Tetrahydronaphthalene.
..
a
Total Extn. Loss after: 48 hr. 96 hr. 48 hr.o Weight per cent 77.3 .. 62.0 .. 61.3 65:l .. 69.5 36.4 5315 7017
..
Volatile Matter 70.0 62.7
Moisture 0.7
0.5
Fixed Carbon 29.2 36.3
TABLE V. DISTRIBUTION OF OXYGENAMONG PRODUCTS
.. 70:7
Ash 0.1 0.5
TABLE IV. PITCHANALYSES(PER C'ENTBY WEIQHT) r
H 6.3 6.5 6.2 6.8 6.4 6.4 6.5 6.4 6.4 6.6 5.6
Ash, Moisture-FreeN 0 87.8 1.8 3.6 1.6 88.4 3.1 1.8 88.6 3.0 2.5 87.9 2.5 3.2 87.7 2.3 1.8 3.2 88.2 88.9 1.6 1.4 89.1 1.8 2.2 1.9 2.7 88.5 89.2 1.6 2.2 84.2 1.7 6.8
C
Oxygen in Original CoalBituminous Brown coal coal oarbon-Present workhydrogenized at Expt. 16 Expt. 17 ated (19) 520° 0. (67) 7 7 35.6 70.8 7 15 46.0 6.6 4 2.8 25 ' 15 15:5 42 39 .. 1.9 10 12 17.9 7%
6i:O
Several significant conclusions may be drawn from the pitch analyses in Table IV. With certain understandable exceptions, the pitches are rather uniform in composition. The high nitrogen content of the pitches from experiments 18 and 19 are probably due to incomplete separation of the nitrogen-containing vehicles on distillation. As was to be expected, addition of hydrogen and removal of inorganic elements as hydrogen sulfide, etc., increased the hydrogen and carbon contents and lowered the carbon-hydrogen ratio. Both the sulfur analyses and the sulfide-like odor of the products show that the total sulfur content was considerably lowered. The organic sulfur was reduced from 0.713 to about 0.4 per cent. The high sulfur content of pitch 22 probably arose from the incomplete removal of the sulfur added originally as p-thiocresol in the vehicle.
Pitch No. 9 16 17 18 19 20 22 28 43 44 Coal analysis
products which contain a large proportion of the original oxygen in phenols and pitch. Since recovery of all the water formed during hydrogenation is difficult, it is likely that the data for oxygen going into water (Table V) are erroneously low. The carbon dioxide value includes all acidic gases and hence is high. In spite of these errors the data in Table V show clearly that mild hydrogenation gives products whose oxygen distribution is different from that of carbonization products.
144 hr.
The two pitches dissolved with greatest difficulty (17 and 32) were obtained by distillation to 245' C.; the other vehicles were removed by distillation to about 215' C. The usual proximate analysis was carried out for pitches 22 and 28 with the results shown below (in per cent) : Pitch No. 22 28
VOL. 29, NO. 8
S
0.5 0.4 0.4 0.3 0.4 0.4 1.6 0.5 0.5 0.4 1.7
The products, excluding the a-naphthylamine and acetamide experiments, had no odor of ammonia. This fact and the nitrogen analyses indicate that the mild conditions used in the present work, although sufficient to eliminate considerable quantities of sulfur and oxygen, failed to remove much nitrogen. Since the recovered hydrocarbon vehicles contained tar bases, some of the coal nitrogen must have been converted into amines. More than half the oxygen was removed during the hydrogenation (Tables IV and V). The distribution of oxygen among the products of two experiments (hydrocarbon vehicles) is shown in Table V. Unlike the carbonization process which converts most of the oxygen into water, hydrogenation under the conditions of the present experiments gave
Appearing in Products as Constituent of: Water (constitution) Carbon dioxide Carbon monoxide Phenols Pitch Residue (or coke) Total
-
-
.. -
73
92
97.1
... 100.0
In no case was i t demonstrated that the vehicle could be quantitatively recovered by distillations although good yields were usually obtained. In a number of experiments the recovered vehicles were extracted with alkali and mineral acid to determine the tar acid and base contents (Table VI). The presence of tar acids and bases in the recovered hydrocarbon vehicles indicates that these products were formed from the coal during the liquefaction. Analysis of the recovered phenol and amine vehicles for tar acids and bases showed that these compounds resist hydrogenation under the present experimental conditions. The low tar acid content of the distillate and the high sulfur content of the pitch (Table IV) indicate that the recovery of p-thiocresol in experiment 22 was poor. The high yield of water and the high tar acid content of the distillate boiling up to 202" C. (experiment 33, Table 111) show that catechol (boiling point 240" C.) was hydrogenated to phenol. The composition of some of the hydrogenation gases is shown in Table VII. Experiments for which gas analyses are not reported yielded gaseous products similar to those of the first four experiments in Table VII. When the composition of the gas varied much from that of these four experiments (tetrahydronaphthalene, naphthalene, tetrahydronaphthalene-naphthalene, and tetrahydronaphthalene-cresol vehicles) the differences could be traced to reactions of the vehicle. When stable vehicles were used, the gases contained methane and ethane as the chief constituents (excluding hydrogen) , the first hydrocarbon predominating. Only negligible amounts of unsaturated hydrocarbons or substances soluble in sulfuric acid were found in the gases. The presence of large amounts of methane in experiments 34, 35, 37, and 39 indicates that this hydrocarbon was formed by hydrogenation of the vehicle. The high ethane contents of the gases from the glycol and acetamide experiments might be expected. The carbonyl groups in the vehicles [abietic acid ( I " ) , stearic acid, and ethyl dihydroabietate] were responsible for the high yields of carbon dioxide and monoxide in experiments 14, 31, and 35. It is interesting to note that carbon monoxide, but not carbon dioxide, was obtained in increased yield when acetamide was the vehicle. There is no obvious reason for the high carbon dioxide content of the gas from the glycol experiment. The high yield of acidic gases in experiment 22 was probably due to the conversion of some of the vehicle (p-thiocresol) into hydrogen sulfide. The acetone- and benzene-insoluble residues were brown or black finely divided solids. It was difficult to identify con-
AUGUST, 1937
INDUSTRIAL AND ENGINEERING CHEMISTRY
as a measure of the fluidity of the products a t room temperature. I n confirmation of previous claims, tetrahydronaphthalene was found to function excellently as a hydrogenation medium. However, the addition of stable polar compounds (toluidine, cresol, quinoline, etc.) to tetrahydronaphthalene gave vehicles which were equally satisfactory or superior. Because coal hydrogenation proceeds better in the presence of acids (24), it is somewhat astonishing that mixtures of amines (otoluidine, quinoline, etc.) with tetrahydronaphthalene gave good results. However, amines (2, 3) such as toluidine and quinoline are known to be effective in dissolving or depolymerizing the coal substance. Several hydrocarbons (tetrahydronaphthalene, methylnaphthalene, and diphenyl) were found to be good vehicles. Methylnaphthalene is especially interesting because, like tetrahydronaphthalene, it gave a product fluid a t room temperature. Why methylnaphthalene is a good dispersion medium is not known. The superiority of methylnaphthalene is in harmony with the observation (IO) that low-temperature tar (containing alkylated aromatic hydrocar-
TABLEVI. DISTILLATION OF CENTRIFUQBD OIL Expt.
No.
DistillaOil tion End Distillate Distilled Point Water Oil Umms * C. CC. Cc.
Vehicle
Tetrahydronaphthalene Tetrahydronaphthalenecresol 15 Tetrahydronaphthalenetoluidine 19 Tetrahydronaphthaleneouinoline 13 Dedahydronaphthalene 17 llethylnaphthalene 22 Tetrahyhronaphthalene thiocresol 0 A t 6 mm.
2 9
127.3a:::{,-_-
Gravity at 60" F. Vol: % Distillate (15.6' C.) Tar Tar Disacids baaes tillate Vehicle 'A.P,I.'A.P.I
'7-,.5}
"*
2.9
0.45
..,
...
10.0
10.0
46.0
12.5
...
...
...
35.6
4.8
1.4
...
11.0 29.3 11.9
27.6
1.3
71 83 79
... ...
0.8
72
9.2
...
13.4
...
Volume CO CH4 0.3 3.6 0.5 5.3
CzHt 3.9 3.4
N2 1.8 0.4
148.5
214
2.0
80
135.5
216
1.0
75
149.3 88.1 139.2
245 207 245
2.3
137.9
220
60.0
... ...
TABLE VII. ANALYSISOF HYDROQENATION GASES Expt. No.
+
Vehicle COr H2S 1.3 2, 8, 12 Tetrahydronaphthalene 1.3 10 27 Xaphthalene 16: 44 Tetrahydronaphthalenenaphthalene 0.7 7, 9 TetrahvdronaDhthalene1.0 11.1 14 22 I 2.4 31 3.6 ste 34 Tetra 4.4 glycol 10.0 35 Ethyl dihydroabietate 37 Tetrahydronaphthalene0.5 nicotine 39 Tetrahydronaphthaleneacetamide 0.0
Per Cent by CnHm 02 HZ 0.1 0 . 5 88.4 0.1 0.3 88.9
0.1
0.5
90.1
0.4
4.4
2.8
1.1
0.0
0.6
0.2 1.0
4.5 5.9
2.0 3.3
1.7 0.3 0.0
0.0
0.5
90.0 77.9
0.2
0.4
88.6
0.1
5.5
2.9
0.2
0.7
84.4
2.0
6.1
3.3
0.0
0.6
0.6 0.8
60.6 67.2
2.9 1.9
10.8 15.4
18.2 4.8
1.9 0.0
0.0 0.0
0.1
83.5
0.3
10.6
4.3
0.7
0.3
0.3
73.0
3.1
10.2
13.0
0.1
943
I
i
TABLEVIII. stituents in these residues with the microscope, but their colloidal nature and the excellent distribution of the catalyst were established. The acetone residue from experiment 17 was analyzed with the following results (per cent by weight) : hydrogen, 3.3; carbon, 60.3; nitrogen, 1.3; oxygen, 0.8; sulfur, 3.7; ash, 30.6. The amounts of hydrogen consumed in the liquefaction experiments were calculated and found to vary from 1.5 to 3.5 grams, Since the possibility of slight leakage cannot be completely excluded, great significance cannot be attached to these values. However, in most cases the calculated hydrogen absorption agrees well with the other experimental data. For example, only 1.5 and 1.9 grams of hydrogen were consumed in the diphenyl ether and diphenylmethane experiments, respectively, whereas considerably more hydrogen was absorbed in other experiments in which extensive liquefaction occurred (Table VIII). The vehicles reacted with hydrogen in the naphthalene, tetrahydronaphthalene-quinoline, tetrahydronaphthalene catechol, tetrahydronaphthalene glycol, and tetrahydronaphthalene-acetamide experiments (2.7, 3.2, 3.2, 3.5, and 3.4 grams of hydrogen absorbed, respectively). In the diphenyl experiment the outlet tubing became clogged, and a gas sample was not collected. A similar experience with diphenyl was reported by Beuschlein, Wright, and Wil, liams (6).
-
-
Discussion of Results Several vehicles were found to be excellent media for the liquefaction of Bruceton coal. These and the other solvents are arranged in Table I11 approximately in the order of decreasing merit, the degree of liquefaction being judged by the amount of residue insoluble in acetone and benzene. Although considerable differences in effectiveness were observed, it was impossible to make exact comparisons or extensive correlations between constitution of vehicle and effectiveness. The amount of ash-free oil obtained on centrifuging the crude product (column E, Table 111) may be taken
MAXIMUMAND FINALPRESSURES OF HYDROGEN ABSORPTION Pressure, Lb./Sq. In. .Maximum Final
H drogen
Expt. No.
Vehicle
2, 8 7, 9
Tetrahydronaphthalene Tetrahydronaphthalenecresol Naphthalene Tetrahydronaphthalenenaphthalene Tetrahydronaphthalene-
2468
775
2.4
2450 2335
765 658
2.2 2.7
2340
705
2.1
N;&&&e-creeol Diphenyl-cresol Meth lnaphthdene TetraEydronaphthaleneoatechol TetrahvdronaDhthalene-
2125 ___. 520
2300 2300 2230
640 695 625
3.2 2.6 2.3 2.8
2440
610
3.2
Axsorbed Gams
10 27 16: 44 19, 32
miinnline
20. 41 24, 42
:.!
~
34
630
3.5
550 810 750
3.4 1.5 1.9
bons) is a better vehicle than high-temperature tar (containing large amounts of naphthalene and other completely aromatic hydrocarbons). Two completely aromatic hydrocarbons, naphthalene and chrysene, were poor vehicles. Diphenylmethane, which is principally aromatic, was also a poor liquefaction medium. Permanent absorption of hydrogen (little or no subsequent dehydrogenation or transference of hydrogen to coal) would explain the failure of these aromatic hydrocarbons to function as good vehicles. It is interesting that, when naphthalene was used with tetrahydronaphthalene or cresol, the liquefaction of coal was better than with naphthalene alone. The saturated hydrocarbons, decahydronaphthalene and n-hexadecane, were very unsatisfactory, probably because of their poor solvent action for coal. The poor results obtained with decahydronaphthalene (as well as in the pinene-tetrahydronaphthalene experiment) may be attributed in part to its low critical temperature (384' C.). The hydrogen-rich compounds (decahydronaphthalene, pinene, abietic acid, and ethyl dihydroabietate), which con-
INDUSTRIAL AND ENGINEERING CHEMISTRY
944
ceivably might have evolved hydrogen during the experiment, proved very unsatisfactory as vehicles. Perhaps higher temperatures or more suitable catalysts would have converted these compounds into hydrogen and (at least for decahydronaphthalene) good liquefaction media. Another hydrogen-rich compound, nicotine, was very effective, but whether this resulted from hydrogen liberation or from its polar properties is not known. Although all the other cresol-hydrocarbon mixtures were good vehicles, the cresol-fluorene mixture was only fair. It is interesting to note that fluorene is somewhat similar in structure to diphenylmethane, which is a poor vehicle. Catechol, an interesting compound because of its depolymerizing action on coal derivatives, was an unsatisfactory vehicle, partly because its conversion into phenol and water required a considerable amount of hydrogen. Several other polar vehicles (p-thiocresol, acetamide, and glycol) were also hydrogenated during the liquefaction experiments.
VOL. 29, NO. 8
vehicles of high molecular weight are superior to those of low molecular weight and the same chemical nature. If a few vehicles are excluded (especially those modified during the experiment), several general correlations can be made between the effectiveness and physical constants. Such generalizations cannot be relied upon completely, for too many variables and unknowns are involved. A satisfactory vehicle should be effective in dissolving the coal substance: the situation is therefore somewhat similar to the extraction of coal, in which polar solvents are known to be superior. However, the extraction of coal in the presence of hydrogen is much more complicated. Other factors, such as resistance to permanent hydrogen absorption, ability to act as hydrogen carrier, etc., are involved. I n the following discussion of the relation between effectiveness as a vehicle and physical constants the constants are given for 20' C. unless otherwise stated. None of the vehicles with low surface tensions (below 30) proved to be effective (Table I). However, the surface tension cannot be taken as a complete measure of the vehicle's merit. For example, OF HYDROGENATED COALDISTILLATE TABLEIX. COMPOSITION diphenylmethane (37.1 a t 30" C.) and naphthaFraction 1 Fraction 2 Fraction 3 Fraction 4 lene (39.0) were not as good as tetrahydroBoiling range, C . To 150 150-225 225-300 300-370 naphthalene, which has a surface tension of 34. Yield cc. 37 (0.75CC.of Hz0) 138 5 275 293 Yield' % 6 18.6 37 39.4 Tetrahydronaphthalene, however, has a dipole Graviiy at 60' ,F. (15.6' C.) (hydrometer), A . P. I. 36.Sa 23 2 15 8 moment of 0.52, while the other two hydrocarSp. gr. (pycnometer) o.822 at 870 F* '41 959a '.OZ9 ai 860 F. bans have zero moments. The greater resist(30.6' C.) (30' C . ) Tar acids 6.0 21 4 15 9 7.4 ance of tetrahydronaphthalene to hydrogenation Tar bases: $ 6.6 3 0 21 0 6 and its hydrogen-carrying capacity are probably Neutral oil constituents. %. Olefins 10 0 8 10 largely responsible for its superiority. Aromatics 33 50.7 70.8 75 Tetrahydronaphthalene (0.52) was the only Unsulfonated oil (paraffins and naphthenes), % 57 49 3 21 2 15 vehicle with a dipole moment under 1.3 to funcParaffin fraction: tion satisfactorily as a vehicle. Some of the worst Refractive index at 85O F. (29.4' C . ) 1.4115 1.4324 1.4558 1 4678 vehicles (cetane, diphenylmethane, naphthalene,' S p r (pycnometer) a t 85' F. 0.740 0 786 0 835 0.858 decahydronaphthalene) have zero dipole moAn8I;e point, D C . 49.3 56 8 71 6 .. ments; other vehicles with low moments (dia Calculated. phenyl ether 1.05, pinene 1.10) were also unsatisfactory. However, one hydrocarbon with a zero moment (diphenyl) was- a fair vehicle. Two fractions (boiling at 185" to 215' and 270" to399' C.) None of the good vehicles has dielectric constants below of a hydrogenated coal distillate furnished by T. E. Warren of 2.5, and only one (diphenyl, 2.62) has a constant below 3.3. the Department of Mines, Ottawa, Canada, were used under However, one vehicle (diphenyl ether, 3.69) with a constant the standard conditions of the present work in the liquefacabove 3.3 was a poor dispersion medium. In most cases the tion of Bruceton coal. The analysis of the hydrogenated best vehicles (cresol 5.5, quinoline 8.9, toluidine 6.0, and xycoal from which these fractions were taken is given in Table lenol 4.8) have high dielectric constants. IX. The lower boiling fraction (185' to 215" C.) was a poor vehicle (75 per cent conversion of pure coal), partly because Acknowledgment of the low critical temperatures of some of its constituents The authors desire to express their sincere appreciation of and the hydrogenation of the vehicle (distillation indicated the generous assistance given by H. M. Cooper, R. F. Aberthat considerable quantities of lower boiling material had nethy, and W. A. Selvig of the Coal Analysis and Miscellanebeen formed). The product obtained with the higher boiling ous Analysis Sections. The gas analyses were carried out fraction (270" to 399' C.) was a gummy tar which, although under the supervision of H. H. Schrenk of the Gas and Dust not fluid a t room temperature, was almost completely disSection, Health Division. Reinhardt Thiessen of the Coal solved (excluding ash, fusain, and catalyst) by benzene in a Constitution Section made the microscopic examinations. Soxhlet extractor. Special thanks are due H. H. Storch and his associates in Technical dimethylnaphthalene (containing 3.3 per cent the Physical Chemistry Section for many helpful suggestions tar acids and 8.9 per cent tar bases) was found to he a fair during the course of the work and the construction of the hyvehicle; 85 per cent of the dry, ash- and fusain-free coal was drogenation equipment. liquefied. The vehicle recovered by distillation to 260' C. contained 4.3per cent tar acids and 7.8 per cent tar bases. Literature Cited Anthracene oil was found to he a good vehicle; over 90 per Am. SOC.Testing Materials, Process 35, Part I, p. 847 (1935). cent of the pure coal was liquefied. The anthracene oil used Aabury, R. S.,IND. ENQ.CHEM.,28, 687 (1936). contained 3.0 per cent tar acids and 4.6 per cent tar bases; Bakes, W. E., Dept. Sci. Ind. Research (Brit.), Fuel Research on distillation, 3 per cent came over up to 200" C. and 68 per Tech. Paper 37 (1933). cent up to 200" C. (9 mm.). The ultimate analysis of the Bermnann, E., and Weizman, A.. Trans.Faradav SOC.,32. 1318 (1936). oil follows (per cent by weight) : carbon, 90.6; hydrogen,6.4; (5) Beuschlein, W. L.,Wright, C. C., and Williams, C. M., IND. nitrogen, 0.8; sulfur, 0.6; oxygen, 1.6. The anthracene oil ENQ. CHEM.,26, 465 (1934). experiment, as well as the ones employing naphthalene, chry(6) Bhatnagar, 8. S.,and Singh, B., J . Indian Chem. SOC.,6 , 263 sene, and the two hydrogenated coal distillates, indicates that (1929). O
AUGUST, 1937 (7)
INDUSTRIAL AND ENGINEERING CHEMISTRY
Boomer, E. H., and Saddington, A. W., Can. J . Research, 12, 825 (1935).
( 8 ) Boomer, E. H., Saddington, A. W., and Edwards, J.. Ibid., 13, Sect. B, 11 (1935). (9) Cork, J. M., Rev. Sci. Instruments. 1, 563 (1930). (10) Dept. Sci. Ind. Research (Brit.), Rept. Fuel Research Board for Year Ended March 31, 1936. (11) Estermann, J., Z. physik. Chem., B1, 134 (1928). (12) Fieldner, A. C., and Davis, J. D., U. 5. Bur. Mines, Monograph 5 (1934). (13) Graham, J. I., and Skinner, D. G . , S. Soe. Chem. Ind., 48, 129T (1929). (14) Graham, J. I., and Skinner, D. G., Proc. 3rd Intern. Conj. Bituminous Coal, 2, 17 (1931). (15) Hem, W., and Schuftan, P., 2. physik. Chem., 101, 269 (1922). (16) Kumler, W. D., and Porter, C. W., J . Am. Chem. SOC.,56, 2549 (1934). (17) LaLande, W. A., Jr., IND.ENQ.CHEM.,26, 678 (1934). (18) Lantsch, W., 2. physik. Chem., B1, 115 (1928). (19) Makray, I. von, Brennstog-Chem., 11, 61 (1930). (20) Marti, F. B., Bull. soc. chim. B e l g . , 39, 590 (1930).
DER ALCHIMIST By Martin Johann Schmidt
Here we see the effect of Shakespeare’s witches upon the alchemists’ activities. Just what rare element the scorpion contains and its effect upon the elixir of life or love philtre being prepared are, of course, unknown. Schmidt was born near Krems, Austria, in 1718, and studied in Vienna and Venice. He spent most of his life in Stein on the Danube, where he became a city official, dying here in 1801 after an active career as a painter and etcher. This is No. 80 in the Berolzheimer series of Alchemical and Historical Reproductions, having been copied from an engraving by Ferdinand Landerer of the original painting.
No. 74 appears on page 166 February issue No. 76 on page 345, March’issue, No. 76 0; page 459, April issue, No. 77 on page, 5 54, iMay issue N o 78 on page 710 June issue, and No. 74 on page 776, July &e.
0
945
(21) Pertierra, J. M., Anales soc. espafi. j i s . quim., 32, 702 (1934). (22) Pertierra, J. M., J . Inst. Fuel, 9, 16 (1935). (23) Petroleum Times, 35, 609 (1936); Chem. Trade J . , 99, 133 (1936). (24) Pier, M., German Patent 597,255 (May 19, 1934). (25) Pott, A., and Broche, H., CXtIckauf, 69, 903 (1933). (26) Schmidt, L. D., Elder, J. L., and Davis, J. D., IND.ENQ. CHEM.,28, 1346 (1936). (27) Thau, A., “Die Schwelung von Braun- und Steinkohle,” p. 36, Halle (Saale), W. Knapp, 1927. (28) Tongue, H., “Design and Construction of High Pressure Chemical Plant,” p. 203, London, Chapman and Hall, Ltd., 1934. (29) Van der Pyl, L. M., Am. Gas Assoc. Proc., 1933, 870. (30) Warren, T. E., and Gilmore, R. E., IND.ENQ.CHEM.. 29, 353 (1937).
RECEIVED February 16, 1937. Presented before the Division of Gas and Fuel Chemistry at the 93rd Meeting of the American Chemical Society, Chapel Hill, N. C., April 12 to 15, 1937. Published by permission of the Direotor, U. 8. Bureau of Mines. (Not subject t o copyright.)
